[mgalusha]

Some ramblings about adding an active crossover to my system.

Back in 2002 I read all of Rod Elliott's articles on actively bi-amping and went so far as to buy a Marchand XM9 crossover but the speakers I had at the time (Vandy 3A Sigs) didn't lend themselves to to easily modifying the XO and I never really did anything with this and eventually traded the Marchand for some other gear.

While cleaning out a file cabinet last week I can across copies of the articles I had printed out 4 years ago and started running things through my head. My current speakers (Meadowlark Blue Heron 2) have a readily accessible XO so that part was not a problem. Re-reading all of Rod's articles he suggests cross over points of about 300Hz and 3600Hz as this is the most critical range for intelligibility. This is a pretty wide range for a midrange driver to cover as it needs to play well for an octave below and above the crossover frequencies. Since the BH2's are first order speakers they need to use high quality drivers to work well and looking up the specs for the midrange (ScanSpeak 12M/4631) showed an Fs of 75Hz and is -3dB at about 175Hz, so the low end appeared like it would be usable. The existing mid to high XO was at 4K, close enough to 3k6 to not worry about it.

A call to my friend Jerry confirmed that he still had a Marchand XM44 crossover I could borrow along with a couple of spare  amps I could experiment with. An order to Mouser for the needed resistors and caps to build the correct high and low pass modules and some very simple modifications to the passive crossover to bypass the woofer section and high pass to the midrange and I connected everything up yesterday morning.

I set the low to mid crossover to 275Hz using Linkwitz-Riley 4th order slopes. I left the internal passive mid/high network in place including the baffle step compensation. I used my R.E. Designs LNPA amps for the mid/high section, these are capable of 150W with a 4R load. I used a Rotel RB1080 for the woofers. This amp is rated to produce 400W with a 4R load. The Rotel has 2.5dB more gain than the LNPA's, so I adjusted the XO accordingly. I wanted the be sure the drivers were receiving equal power as all I wanted to change was the crossover frequency and the slope. I didn't want to boost or cut either the bass or mid/high levels.

I had high hopes for the sound and thus far the results have exceeded my expectations. Bass is amazingly tight since the amp is directly coupled to the drivers and doesn't have a large inductor in series to cause problems. This I expected but I did have some concerns about crossing over the midrange so low but with a 4th order slope vs 1st order it's actually seeing less low frequency energy than with the original crossover at 2KHz. The improvement in the mids and highs was even more surprising than the improvements to the bass. Much greater clarity in both the midrange and the highs. I liked the tweeter (ScanSpeak R2904/7000) before but it's substantially better in this configuration. Even though it was brought in at 4KHz it seems like it was still receiving excess low frequency information due to the 1st order slope.

I need to find the rest of my measurement gear and take some in room measurements but since all the drivers are still seeing the same levels, the response should be reasonably flat. I certainly sounds good and I couldn't be happier for a first try at this, especially with a very minimal investment. Of course using an active crossover is nothing new but until I heard the drastic transformation in my system first hand I didn't expect the improvement to be that significant. I keep trying to find something that doesn't sound right or that indicates a problem but after a great deal of listening the past two days I haven't heard anything I didn't like. The use of a digital crossover would give me even more flexibility but I have a pretty good vinyl rig and I don't really like the idea of doing an A/D - D/A conversion to play records, so I may end up staying with an analog XO. One thing is certain, I can't go back to my previous setup, which was pretty damn good but this has moved it to an entirely new level.

Granted this is not a trivial exercise and is certainly not for everyone, it has turned out great and was well worth the time and effort.

Thanks for taking the time to read my ramblings.

mike

[woodsyi]

Mike,

With all due respect to Dan's amp, try some tubes on top and see what you hear.  aa  Also, UcD or ICE amps are great with woofers. When you get it right, the CLARITY is amazing.  Isn't it? 

[andyr]

Some ramblings about adding an active crossover to my system.

Back in 2002 I read all of Rod Elliott's articles on actively bi-amping ...

My current speakers (Meadowlark Blue Heron 2) have a readily accessible XO so that part was not a problem. Re-reading all of Rod's articles he suggests cross over points of about 300Hz and 3600Hz as this is the most critical range for intelligibility. This is a pretty wide range for a midrange driver to cover as it needs to play well for an octave below and above the crossover frequencies. Since the BH2's are first order speakers they need to use high quality drivers to work well and looking up the specs for the midrange (ScanSpeak 12M/4631) showed an Fs of 75Hz and is -3dB at about 175Hz, so the low end appeared like it would be usable. The existing mid to high XO was at 4K, close enough to 3k6 to not worry about it.

I set the low to mid crossover to 275Hz using Linkwitz-Riley 4th order slopes. I left the internal passive mid/high network in place including the baffle step compensation.

I had high hopes for the sound and thus far the results have exceeded my expectations. Bass is amazingly tight since the amp is directly coupled to the drivers and doesn't have a large inductor in series to cause problems.

Granted this is not a trivial exercise and is certainly not for everyone, it has turned out great and was well worth the time and effort.

mike

Hi mike,

Glad to see you like active!    I converted my Maggie IIIAs to 3-way active about 10 years ago and have never looked back!    As you say, removing those large inductors enables the amp to grab hold of the bass driver with an iron fist!!

Just a few comments on what you said:

1.  I also read all of Rod's articles and in fact use his crossover PCBs.  Yes, he says the theoretical "best crossover points" are 300Hz & 3600Hz but in the real world you have to stick with what your drivers are capable of ... my IIIa crossovers for instance are 300Hz and 2600Hz (because my mid-range driver won't go up that high!).

2.  First up, I built Rod's 24dB L-R XOs in a 3-way format.  But then I built a second pair using the stock Maggie IIIa slopes (18dB bass LP/12dB mid HP & 12dB mid LP/12dB ribbon HP) ... and these sound much better.

I suggest it's possible your Meadowlark BH-2s would sound "even better )" if you changed the bass/mid crossover slopes to 6dB (some Marchand models let you do this and Rod's XO PCB certainly does) AND stuck to the stock XO frequency.

3.  Whatever sonic benefit you can hear with an active bass/mid XO, you will get more of with an active mid/tweeter XO ... so why not go 3-way active!!??  )  If you do this, you would have to leave in place the passive baffle-step compensation ... plus, I suggest, any driver impedance compensation components (IMO, they are not part of the XO network!).

Regards,

Andy

[mgalusha]

With all due respect to Dan's amp, try some tubes on top and see what you hear.  aa  Also, UcD or ICE amps are great with woofers.

Woodsyi,

I might give that a try one of these days but I have had several very nice tube amps in my system with these speakers, admittedly running full range, and in the end I preferred the LNPA's. They are very good sounding amps IMO.

One of the first things I thought of was a UcD for the bass and I may still go that way but since I was able to borrow both the Rotel and a Belles 350A I took advantage of the opportunity.

mike

[mgalusha]

Hi Andy,

Thanks for the feedback.

1.  I also read all of Rod's articles and in fact use his crossover PCBs.  Yes, he says the theoretical "best crossover points" are 300Hz & 3600Hz but in the real world you have to stick with what your drivers are capable of ... my IIIa crossovers for instance are 300Hz and 2600Hz (because my mid-range driver won't go up that high!).

I was fortunate on the midrange in that it's nearly flat to 6KHz, at least on axis and was capable of going to 300 without problems.

2.  First up, I built Rod's 24dB L-R XOs in a 3-way format.  But then I built a second pair using the stock Maggie IIIa slopes (18dB bass LP/12dB mid HP & 12dB mid LP/12dB ribbon HP) ... and these sound much better.

I suggest it's possible your Meadowlark BH-2s would sound "even better Smile)" if you changed the bass/mid crossover slopes to 6dB (some Marchand models let you do this and Rod's XO PCB certainly does) AND stuck to the stock XO frequency.

This may be but I think part of the reason I'm getting so much better clarity is because the woofers are no longer reproducing the midrange, the stock bass/mid xo point was 2KHz. The nice part of using the XM44 is that it uses plug in modules so I can experiment with different frequencies and slopes by just plugging in a new module. I have some spares for just this purpose.

3.  Whatever sonic benefit you can hear with an active bass/mid XO, you will get more of with an active mid/tweeter XO ... so why not go 3-way active!!??  Smile)  If you do this, you would have to leave in place the passive baffle-step compensation ... plus, I suggest, any driver impedance compensation components (IMO, they are not part of the XO network!).

Oh I'm already considering it, just taking things one step at a time. The XM44 will accommodate baffle step compensation but obviously can't do anything about impedance compensation. I would like to get the LP inductor off of the midrange as I'm sure that will make a nice improvement. There is of course the matter of yet another amp, so while I may give this a go, it will be a while before I get there.

mike

[Dan Banquer]

Did you find any changes in room /loudspeaker issues? Or in your initial findings did you find that with a partial  active crossover the room loudspeaker issues were increased or decreased. Did the typical room gain that many speakers have in the bass either, lessen, increase or no change at all?
             d.b.

[mgalusha]

Did you find any changes in room /loudspeaker issues? Or in your initial findings did you find that with a partial  active crossover the room loudspeaker issues were increased or decreased. Did the typical room gain that many speakers have in the bass either, lessen, increase or no change at all?

Dan,

I haven't had a chance to make any in room measurements yet but so far I don't think my room/speaker interactions are much different. My current listening room has quite a bit of acoustic treatment and is a very nice, if somewhat small, listening space and doesn't seem to have a lot of problems.

I do notice quite a bit tighter bass but I'm pretty sure this is due to the amp being directly coupled to the woofers and not having a large series inductor.

mike

[Dan Banquer]

If you've got the equipment or can get your hands on the equipment try  pink noise with a real time analyzer and compare room gain at low frequencies with and without the active crossover. My bet is that there is a difference, and the room gain is lower.
                d.b.

[mgalusha]

Over the holiday I had a chance to make an in room measurement. While not flat it's not too bad for a system that's not using any type of EQ such as a TacT. Since the XM44 is on loan I'm thinking of going with a DCX2496 so that I can apply some (light) EQ to smooth out the bass a bit. My hacked RS SPL meter I used for measuring is supposed to be pretty flat to about 10Hz but the HF response is almost non existent, so the drop above 8K is from the mic, not the room.

[andyr]

If you've got the equipment or can get your hands on the equipment try  pink noise with a real time analyzer and compare room gain at low frequencies with and without the active crossover. My bet is that there is a difference, and the room gain is lower.
                d.b.

Hi Dan,

As a guy that runs his Maggie IIIas 3-way active, I'm interested to understand why you think room gain will be lower (at LFs) with an active setup?  It's not something I thought about when I changed from upgraded passive XOs to active, so I can't say "yeah" or "nay"!!   

Regards,

Andy

[JLM]

Mike,

Thanks for sharing.  But might I suggest that in effect you've only created a passive 2-way with powered subwoofers.  You're still only half way home to enjoying the full benefits of going fully active.

Years ago I had the chance to compare $800/pair Paradigm Studio 20s (typical 2-way standmounts) against $1600/pair Paradigm Active 20s (pretty much a straight passive to active conversion of the same design).  The Actives blew the Studios away, not even close.  Much more dynamic, ruler flat response, and incredible bass output.  They both had the same Paradigm family sound, but a fairer comparison would have been against the Studio 100s ($2000 full sized floorstanders).  Other customers gawked in disbelief at the bass and dynamics coming from standmounts.  Their performance was amazing, the 2nd and last audio epiphany I've had.

From that day I was convinced that active is the only way to go for any serious speakers or audio system.  I might conscend to a two-way using a super tweeter with protective cap (like the DeCapo 3A).  Too bad that most audiophiles haven't heard actives or accept accounts like these.  And unfortunately most active speakers are designed for highly analytical or high output professional use.  IMO the money spent on expensive passive speakers is mostly wasted. 

BTW the active advantages also explains why us owners of single driver speakers aren't completely out in left field. 

[mgalusha]

Thanks for sharing.  But might I suggest that in effect you've only created a passive 2-way with powered subwoofers.  You're still only half way home to enjoying the full benefits of going fully active.

I partially agree. Since the woofers are still running up to 275Hz they cover a lot more ground than subwoofers but I agree, getting rid of the passive XO between the mid and tweeter should be very beneficial. It will come, just a matter of getting some additional amplification in place for the tweeters. It all takes time, baby steps for now.

mike

[Occam]

Mike,

I think your approach is just swell!!! And when you've perfected it, I'm gonna steal it for my Alons.
This division of labor and resources allows the use of an amp on the bottom end the is appropriate for controlling the reactive load of the woofer, and possibly, even the use of a low powered 'hard' amp on the mid/tweet head. By 'hard' I mean an amp with an output impedance approching that of the driver/crossover load. These are usually no-low feedback designs. It may have a pissant dampning factor but on a well behaved mid/tweet head unit, dang!, it sure can sound purty.

From each according to their abilities,
To each, according to their needs.

[ctviggen]

Mike,

Thanks for sharing.  But might I suggest that in effect you've only created a passive 2-way with powered subwoofers.  You're still only half way home to enjoying the full benefits of going fully active.

Years ago I had the chance to compare $800/pair Paradigm Studio 20s (typical 2-way standmounts) against $1600/pair Paradigm Active 20s (pretty much a straight passive to active conversion of the same design).  The Actives blew the Studios away, not even close.  Much more dynamic, ruler flat response, and incredible bass output.  They both had the same Paradigm family sound, but a fairer comparison would have been against the Studio 100s ($2000 full sized floorstanders).  Other customers gawked in disbelief at the bass and dynamics coming from standmounts.  Their performance was amazing, the 2nd and last audio epiphany I've had.

From that day I was convinced that active is the only way to go for any serious speakers or audio system.  I might conscend to a two-way using a super tweeter with protective cap (like the DeCapo 3A).  Too bad that most audiophiles haven't heard actives or accept accounts like these.  And unfortunately most active speakers are designed for highly analytical or high output professional use.  IMO the money spent on expensive passive speakers is mostly wasted. 

BTW the active advantages also explains why us owners of single driver speakers aren't completely out in left field. 

I also think that going completely active is the way to go.  However, it's also very expensive.  You need active crossovers and more amplifiers (one for each crossover region).  Then, if you want to design the active crossover yourself, you need time to fiddle with the crossover.  Every time I consider just biamping my RM40s, for instance, I have to consider the cost of an extra set of amplifiers and something to equalize the sensitivities of the amps.  Then, going completely active is another step (or several steps) beyond that.  It's simply too time consuming and expensive for me. 

One such completely active design I've been examining is the Linkwitz Orions.  He's already done most of the work for you there, though.  Still, the cost of amps for that design could be high, if you're a "better amp = better music" type of person.

[Dan Banquer]

Over the holiday I had a chance to make an in room measurement. While not flat it's not too bad for a system that's not using any type of EQ such as a TacT. Since the XM44 is on loan I'm thinking of going with a DCX2496 so that I can apply some (light) EQ to smooth out the bass a bit. My hacked RS SPL meter I used for measuring is supposed to be pretty flat to about 10Hz but the HF response is almost non existent, so the drop above 8K is from the mic, not the room.

Hi Mike:
Looking at the response you supplied it appears that you have a dip at 60 Hz and a spike at 120 Hz. Also dips at about 180 Hz and 360 Hz. As we go up in frequency this initially appears to be harmonically related. It's things like this that make me wonder if getting a grip on the dip at 60 Hz and spike at 120 Hz just might make the rest of response smooth out.
Are you running a live end dead end? If not; and you are able to, I highly recommend it. It does wonders for midrange on up in smoothing out a response.
                  d.b.

[andyr]

I also think that going completely active is the way to go.  ... have to consider the cost of ... something to equalize the sensitivities of the amps.

Good active XOs have gain controls for each band, so you can compensate for different driver sensitivities and amplifier gains.

Regards,

Andy

[JLM]

Here's a link to add inspiration (to go all the way):

http://www.tnt-audio.com/casse/active_speakers_intro2_e.html